Observations made studying anti-tumor immune responses in patients with metastatic melanoma have reshaped critical questions to be addressed mechanistically in preclinical models. In many melanoma patients, and in several animal studies, the rate limiting step in the anti-tumor immune response appear to be at the level of immune resistance at the effector phase in the tumor microenvironment. In the previous funding period we have made significant advances in understanding these mechanisms of resistance and how to overcome them. However, this new perspective on anti-tumor immunity has generated a critical new question: if tumor antigen-specific T cells frequently can be productively activated without any external intervention, then which innate immune factors are facilitating proper T cell differentiation, in the absence of an obvious infectious pathogen? In theory this should occur only with difficulty, as there are no exogenous TLR ligands to costimulate antigen-presenting cell (ARC) populations in the context of a growing tumor. Our preliminary data have indicated a role for host type I interferons (IFNs) and also IL-1 in this innate immune "awareness" that leads to productive tumor antigen cross-presentation to CD8+ T cells in vivo. In the first specific aim of this proposal, we will identify and characterize the ARC populations mediating apontaneous cross-priming of tumor antigen-specific CD8+ T cells in vivo. The rationale is that we need to understand what occurs normally in order to determine where the blocks are in this process in specific mutant mice. In the second specific aim, we will determine the mechanism by which host type I IFNs facilitate tumor antigen cross-presentation in vivo. In the third specific aim, we will elucidate the role of host IL-1 and its relationship to type I IFNs in productive cross-priming in vivo. These studies will take advantage of several unique reagents and model systems, including a TCR tetramer to recognize peptide/class complexes, conditional IFN-a/bR knockout mice, and a genetic tumor model encoding a model tumor antigen. Understanding the factors and processes that mediate successful T cell priming to tumor antigens when it does occur should point toward new strategies to induce better cross-priming when it does not occur spontaneously.